This is great! Structured extraction is exactly what Paperless has been missing. The schema-based JSON output + simple Docker setup is a big win. I'll probably

This is great! Structured extraction is exactly what Paperless has been missing. The schema-based JSON output + simple Docker setup is a big win. I'll probably try it on some ugly real-world receipts.